Poster #109 - Object kinds interfere with extending novel adjectives: Evidence from young children and adults

• In Event: 3-053 - Poster Session 10
  In Poster Session: PS 10 Section - Language, Communication

Integrative Statement

Across languages, the acquisition of adjectives tends to lag substantially behind the acquisition of nouns and verbs (Fenson et al., 1994; Jackson-Maldonado et al., 1993; Tribushinina et al., 2014). One central challenge for adjective acquisition may be their breadth of application. Three-year-old children who can extend an adjective successfully within a particular category (from one red car to another), nevertheless struggle to extend the same adjective to a member of a different category (to a red airplane) (Klibanoff & Waxman, 2000; Mintz & Gleitman, 2002).

We propose that these difficulties stem from children’s tendency to focus on object categories, rather than object properties (c.f. Mintz, 2005). We test this idea, first with children and then with adults, by constructing exemplars that can be flexibly construed as representing either objects or blobs of stuff. If learners’ adjective extension is impaired by object categories, then they should show improved adjective extension when they construe the same exemplar as a blob of stuff.

In Experiment 1, 39 children (mean age = 35.6 months) were randomly assigned to either the Stuff or Object condition. Participants in each condition saw the same images: the only difference was in how the experimenter described them: as either “blobs of stuff” or “pictures of things” (Figure 1). Then, in each of 12 adjective extension trials, the experimenter pointed to the target, saying “This is a blickish one” and then revealed the two test stimuli, asking “Can you find another one that’s blickish?” The test stimuli were unambiguous object images, and only one matched the target exemplar in color.

As predicted, children more successfully selected the color-matched test image in the Stuff (M=.94) than in the Object condition (M=.75), X^2(1)=3.95, p=.047. Thus, even when extending adjectives to objects, children performed better—near ceiling—when they construed the target exemplar as a blob of stuff, not an object.

In Experiment 2, we used a parallel design to test whether this tendency persists into adulthood. Seventy-nine adults were assigned to the Stuff or Object condition and performed 6 adjective extensions, featuring the images from Experiment 1 that can be flexibly construed. As before, only one test exemplar matched the target exemplar’s color.

As expected, adults in both conditions were accurate in their adjective extensions (M=.98); however, different patterns emerged in their reaction times. A growth-curve model revealed both linear and quadratic effects of trial, ps<.001, and a marginal effect of condition: adults made extensions more quickly in the Stuff Condition (M = 1119ms) than the Object condition (M=1293), p=.065. Moreover, these effects were qualified by a significant interaction between condition and quadratic time, p=.013, suggesting that construing exemplars as stuff facilitated learning across trials (Figure 2).

These results reveal remarkable continuity across development. Both children and adults extended adjectives referring to an exemplar’s color more readily if they construed that exemplar as a blob of stuff, rather than as an object. This outcome suggests that the presence of object kind interferes with adjective learning, posing a substantial obstacle for adjective acquisition.

Authors

• Alexander LaTourrette, Northwestern University
  Presenting Author

• Sandra Waxman, Northwestern University
  Non-Presenting Author